Methods for the synthesis of speech are known wherein different phonemes are obtained by combining sinusoidal oscillations of respective frequencies and respective amplitudes. Apparatuses implementing such methods are complex and require analog generators with complicated tuning.
Other devices are known which utilize large memories stored on magnetic disks. The vocabularies of such devices are nevertheless limited.